COVID-19: cases and deaths

Abstract

This report analyzes cases and deaths attributed to COVID-19 since January, 2020.

Resources

• Data from Our World in Data.
• plotly for data visualization.

Modules

import pandas as pd
import plotly.express as px
import plotly.graph_objects as go
from plotly.subplots import make_subplots

Dataset

Variables:

• iso_code: ISO 3166-1 alpha-3 – three-letter country codes.
• continent: continent of the country.
• location: name of the state or federal entity.
• date: date of the observation (year-month).
• cases: new confirmed cases of COVID-19 per 1M people.
• deaths: new deaths attributed to COVID-19 per 1M people.

Data Analysis

df = pd.read_csv('./cases_deaths1.csv', header = 0)

Table 1. Cases and deaths from COVID-19.

df.tail(10)

	iso_code	continent	location	date	cases	deaths
140	ITA	Europe	Italy	21-Dec	18174.3	59.2
141	JPN	Asia	Japan	21-Dec	44.9	0.3
142	GBR	Europe	United Kingdom	21-Dec	39855.7	53.7
143	USA	North America	United States	21-Dec	18599.5	134.8
144	ARG	South America	Argentina	22-Jan	24968.5	14.0
145	AUS	Oceania	Australia	22-Jan	43586.3	12.6
146	ITA	Europe	Italy	22-Jan	33626.8	46.2
147	JPN	Asia	Japan	22-Jan	604.6	0.2
148	GBR	Europe	United Kingdom	22-Jan	29878.9	40.0
149	USA	North America	United States	22-Jan	27966.9	60.2

Cases vs deaths in different countries

px.scatter(data_frame = df,
           x = 'cases',
           y= 'deaths',
           size = 'deaths',
           color = 'location',
           title = 'Cases per million vs deaths per million',
           labels = {'cases': 'cases per million', 'deaths': 'deaths per million'},
           # Scales
           range_x = [-1000, 50e+03],
           range_y = [-100, 500],          
           # Hover over points
           hover_name = 'location',
           text = "iso_code",
           # Animation based on date
           animation_frame = 'date',
           size_max = 100)

Figure. Animated pairplot: cases per million vs deaths per million attributed to COVID-19 in different countries. Circle size based on deaths per million. Color key for countries.

It is observed how the plot started with cases lower than 20k but many deaths attributed to COVID-19. By the end of 2021 and begining of 2022, values are below 100 and in some countries near 0 in y-axis showing the high contagiousness of the Omicron variant circulating and causing many cases and the good effectiveness of the vaccines reducing significantly the deaths. It is also noticed that COVID-19 arrived first in Europe or North America countries. Besides, the countries in opposite hemispheres show different behaviors in the coldest or warmest months. For example, Argentina shows the highest number of cases in May - June 2021, while northern countries such as United Kingdom or the United States show the highest in December 2020 and January 2021.

Cases per million

px.area(data_frame = df,
        x = 'date',
        y= 'cases',
        facet_col="location", 
        facet_col_wrap=2,
        color = 'location',
        title = 'Cases per million')

Figure. Faceted area plot: date vs cases per million due to COVID-19 in different countries.

This plot shows the different waves of COVID-19 through time.

Deaths per million

px.area(data_frame = df,
        x = 'date',
        y= 'deaths',
        facet_col="location", 
        facet_col_wrap=2,
        color = 'location',
        title = 'Deaths per million')

Figure. Faceted area plot: date vs deaths per million attributed to COVID-19 in different countries.

Here we could observe the effectiveness of vaccines in the last period of time.

Argentina

Table 2. Data from Argentina.

data_ARG = df.where(df.location == 'Argentina').dropna()
data_ARG.tail(10)

	iso_code	continent	location	date	cases	deaths
90	ARG	South America	Argentina	21-Apr	13782.1	175.6
96	ARG	South America	Argentina	21-May	17638.6	312.0
102	ARG	South America	Argentina	21-Jun	15098.7	355.5
108	ARG	South America	Argentina	21-Jul	10073.1	250.3
114	ARG	South America	Argentina	21-Aug	5610.2	133.6
120	ARG	South America	Argentina	21-Sep	2019.8	11.6
126	ARG	South America	Argentina	21-Oct	699.6	16.9
132	ARG	South America	Argentina	21-Nov	919.6	14.0
138	ARG	South America	Argentina	21-Dec	7096.9	12.7
144	ARG	South America	Argentina	22-Jan	24968.5	14.0

# Figure with secondary y-axis
fig = make_subplots(specs=[[{"secondary_y": True}]])

# Traces
fig.add_trace(
    go.Bar(x = data_ARG['date'],
            y= data_ARG['cases'],
           name="cases per million"),
    secondary_y=False,
)

fig.add_trace(
    go.Line(x = data_ARG['date'],
            y= data_ARG['deaths'], 
            name="deaths per million"),
    secondary_y=True,
)

# Title
fig.update_layout(
    title_text="Argentina"
)

# x-axis title
fig.update_xaxes(title_text="date")

# y-axes titles
fig.update_yaxes(title_text="cases per million", secondary_y=False)
fig.update_yaxes(title_text="deaths per million", secondary_y=True)

fig.show()

Figure. Multiplot: date vs cases per million and deaths per million in Argentina.
LINK to plotly image

Here we observe the dates when different COVID-19 waves happened in Argentina. First one in October 2020 and second one in May - June 2021, both with high number of deaths. The third wave is already happening, in January 2022, but it is well observed the low number of deaths in this case thanks to vaccines effectiveness.

px.area(data_frame = data_ARG,
        x = 'date',
        y= 'deaths',
        labels = {'deaths': 'deaths per million'},
        title = 'Deaths per million in Argentina')

Figure. Area plot: date vs deaths per million in Argentina.

px.scatter(data_frame = data_ARG,
           x = 'date',
           y= 'cases',
           size = 'deaths',
           color = 'deaths',
           title = 'Cases per million and death per million in Argentina',
           labels = {'cases': 'cases per million', 'deaths': 'deaths per million'},
           # Hover over points
           hover_name = 'deaths',
           text = 'deaths',
           color_continuous_scale = 'sunset',
           size_max = 100)

Figure. Pairplot: date vs cases per million in Argentina. Circle size and color based on deaths per million and text inside indicates its value.